Abstract/ Project Summary We propose to identify modifiable factors associated with breast cancer risk and survival and to examine underlying mechanistic pathways. We will utilize an integrative molecular approach that incorporates metabolomic profiles, tumor subtypes, and tissue gene expression data to examine critical, unanswered questions with respect to non-hormonal mechanisms of breast carcinogenesis. The central themes of this project are metabolism, energy balance, and insulin resistance. Measuring plasma metabolites (i.e., final products of the genome), we propose to examine specific branched-chain amino acids valine, leucine, isoleucine as well as metabolomic signatures that may be important in breast carcinogenesis. In addition, we will determine if metabolomic signatures associated with higher breast cancer risk differ between healthy African-American women and healthy Caucasian women. Energy balance and insulin resistance may affect metabolic pathways by inactivating AMPK, thus activating the mTOR pathway and increasing breast cell proliferation. Mutations in PIK3CA or loss of PTEN are present in 50-75% of breast tumors, leading to constitutively active mTOR. However, activation of the PI3K pathway by metabolic events in tumors with an otherwise normally regulated pathway may provide alternative means to encourage breast cancer development and progression. Specifically, we will consider metformin, a dietary diabetes prevention score, and different forms of physical activity including strength training. To better understand biological mechanisms and identify potential preventive targets, we will examine the role of metabolism, energy balance, and insulin resistance with patterns of gene expression in paired tumor/normal tissues. In addition, we will examine exposures by intrinsic subtype (luminal A, luminal B, HER2 type, and basal-like). The mechanisms linking the insulin resistance pathway to breast cancer also apply to risk of metastasis, distant recurrence, and breast cancer mortality after diagnosis. Our Project capitalizes on unique aspects of our cohort, including long-term follow-up, two measures of plasma metabolites 10 years apart, archived tissue specimens, and evaluation of post-diagnostic exposures, controlling for pre-diagnostic behaviors. Findings will substantially enhance our understanding of the etiology of breast cancer, its specific subtypes, and survival. Identifying modifiable factors to improve breast cancer survival has potential translational implications through changes in clinical recommendations. Together with Projects 2, 3, and 4, our aims also will identify etiology and prevention for 3 of the top 5 causes of cancer death.